Skip to content

SANJAI-s0/IRS_genai

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
.idea
.idea
 
 
.vscode
.vscode
 
 
.gitignore
.gitignore
 
 
IRS.py
IRS.py
 
 
LICENSE
LICENSE
 
 
README.md
README.md
 
 
irsai.py
irsai.py
 
 
logo.png
logo.png
 
 
requirements.txt
requirements.txt
 
 

Repository files navigation

🔎 Information Retrieval System using Gemini API (Combined README)

This document combines the README files for both the "Heavier Version" (IRS.py) and the "Lighter Version" (irsai.py) of the Information Retrieval System. This allows for a comprehensive overview of both implementations within a single file.

🚀 README_IRS.md (for IRS.py - Heavier Version)

🔎 Information Retrieval System using Gemini API (Heavier Version)

This is a comprehensive Information Retrieval System built with Python, Streamlit, and Google's Gemini API. Users can upload a variety of document types (PDF, DOCX, TXT, PPTX, images, Excel, CSV) and interact with the content through context-aware answers, summaries, keyword searches, and multimedia exports. This heavier version includes advanced UI/UX features, detailed error handling, and extensive functionality.

📌 Use Case

  • Upload PDF, DOCX, TXT, PPTX, JPG/PNG, XLSX, CSV files (up to 200MB)

  • Ask natural language questions about uploaded content

  • Generate full or topic-based summaries

  • Export responses and summaries as PDFs or text

  • Convert outputs to text-to-speech audio

  • Search for keywords within documents

  • Export complete chat history

  • Common use cases include:

  • 📘 Academic research or literature review

  • 🏫 Educational queries from diverse study materials

  • 📄 Legal/medical/business document analysis

  • 🔉 Audio-based document briefing for accessibility

⚙️ Installation

Follow these steps to set up and run the Information Retrieval System locally:

  1. Clone the Repository
git clone https://github.com/SANJAI-s0/IRS_genai.git
cd IRS_genai
  1. Create and Activate a Virtual Environment (Recommended)
# For macOS/Linux
python3 -m venv venv
source venv/bin/activate

# For Windows
python -m venv venv
venv\Scripts\activate
  1. Install Dependencies
pip install -r requirements.txt
  1. Set Up Environment Variables Create a .env file in the root directory and add your Gemini API key and PostgreSQL database credentials:
GEMINI_API_KEY=your_actual_gemini_api_key
DB_USERNAME=your_db_username
DB_PASSWORD=your_db_password
DB_HOST=your_db_host
DB_PORT=your_db_port
DB_NAME=your_db_name
  1. Run the Streamlit App
streamlit run IRS.py

Your app will be available at http://localhost:8501.

📦 Requirements

  • Python 3.8 or higher

  • Gemini API key (from Google AI Studio)

  • PostgreSQL database

  • Internet connection

requirements.txt

streamlit==1.37.0
python-dotenv==1.0.1
google-generativeai==0.5.4
PyPDF2==3.0.1
pyttsx3==2.90
python-docx==1.1.0
pandas==2.2.2
Pillow==10.4.0
fpdf==1.7.2
psycopg2-binary==2.9.9
SQLAlchemy==2.0.31
unicodedata2==15.1.0
mimetypes==1.2.0

🔐 .env & .gitignore

.env

GEMINI_API_KEY=your_actual_gemini_api_key
DB_USERNAME=your_db_username
DB_PASSWORD=your_db_password
DB_HOST=your_db_host
DB_PORT=your_db_port
DB_NAME=your_db_name

.gitignore

.env
__pycache__/
*.pyc
*.mp3
*.wav
*.ogg

✅ This ensures secret keys and compiled files are not pushed to GitHub.

🛠 Tech Stack

Tool/Library Purpose
Python Core backend logic
Streamlit Web-based UI framework
PyPDF2 PDF text extraction
python-docx Word document parsing
google-generativeai Gemini LLM API access
python-dotenv Environment variable management
pyttsx3 Text-to-speech conversion
fpdf Exporting summaries as PDFs
pandas Excel/CSV processing
Pillow Image handling
psycopg2 PostgreSQL database integration
SQLAlchemy Database ORM
unicodedata2 Text sanitization
mimetypes File type detection

📝 License MIT License — use freely with attribution

🧠 How It Works

  • 📁 Users upload multiple file types (PDF, DOCX, TXT, etc.).

  • 📄 Text is extracted and combined into a unified context with progress tracking.

  • 👤 Users can:

  • Ask natural language questions

  • Request summaries (full or topic-based)

  • Search for keywords

  • Download responses as PDF, text, or audio

  • View and export chat history

  • 🧠 Prompts are constructed using document context and sent to Gemini 2.0 Flash.

  • 📤 Responses are displayed with advanced UI styling and error handling.

  • 📥 Exports include detailed formatting and database storage.

📬 Contact

Built by Sanjai For suggestions or contributions, open an issue or pull request on GitHub: https://github.com/SANJAI-s0/IRS_genai/issues

🚀 Notes for Heavier Version

  • Features: Includes progress bars, tooltips, extensive CSS customization, and robust error handling.

  • Database: Stores chat history in PostgreSQL with SSL support.

  • Limitations: Heavier resource usage due to detailed UI and database operations.

  • Recommendation: Use for production environments with large datasets or complex workflows.

💡 README_irsai.md (for irsai.py - Lighter Version)

🔎 Information Retrieval System using Gemini API (Lighter Version)

This is a streamlined Information Retrieval System built with Python, Streamlit, and Google's Gemini API. Users can upload PDF, DOCX, TXT, Excel, or CSV files and interact with the content through context-aware answers, summaries, and multimedia exports. This lighter version focuses on simplicity, reduced resource usage, and essential functionality.

📌 Use Case

  • Upload PDF, DOCX, TXT, PPTX, JPG/PNG, XLSX, CSV files (up to 200MB)

  • Ask natural language questions about uploaded content

  • Generate full or topic-based summaries

  • Export responses and summaries as PDFs or text

  • Convert outputs to text-to-speech audio

  • Search for keywords within documents

  • Export complete chat history

  • Common use cases include:

  • 📘 Academic research or literature review

  • 🏫 Educational queries from diverse study materials

  • 📄 Legal/medical/business document analysis

  • 🔉 Audio-based document briefing for accessibility

⚙️ Installation

Follow these steps to set up and run the Information Retrieval System locally:

  1. Clone the Repository
git clone https://github.com/SANJAI-s0/IRS_genai.git
cd IRS_genai
  1. Create and Activate a Virtual Environment (Recommended)
# For macOS/Linux
python3 -m venv venv
source venv/bin/activate

# For Windows
python -m venv venv
venv\Scripts\activate
  1. Install Dependencies
pip install -r requirements.txt
  1. Set Up Environment Variables Create a .env file in the root directory and add your Gemini API key and PostgreSQL database credentials:
GEMINI_API_KEY=your_actual_gemini_api_key
DB_USERNAME=your_db_username
DB_PASSWORD=your_db_password
DB_HOST=your_db_host
DB_PORT=your_db_port
DB_NAME=your_db_name
  1. Run the Streamlit App
streamlit run irsai.py

Your app will be available at http://localhost:8501.

📦 Requirements

  • Python 3.8 or higher

  • Gemini API key (from Google AI Studio)

  • PostgreSQL database

  • Internet connection

requirements.txt

streamlit==1.37.0
python-dotenv==1.0.1
google-generativeai==0.5.4
PyPDF2==3.0.1
pyttsx3==2.90
python-docx==1.1.0
pandas==2.2.2
fpdf==1.7.2
psycopg2-binary==2.9.9
SQLAlchemy==2.0.31

🔐 .env & .gitignore

.env

GEMINI_API_KEY=your_actual_gemini_api_key
DB_USERNAME=your_db_username
DB_PASSWORD=your_db_password
DB_HOST=your_db_host
DB_PORT=your_db_port
DB_NAME=your_db_name

.gitignore

.env
__pycache__/
*.pyc
*.mp3
*.wav
*.ogg

✅ This ensures secret keys and compiled files are not pushed to GitHub.

🛠 Tech Stack

Tool/Library Purpose
Python Core backend logic
Streamlit Web-based UI framework
PyPDF2 PDF text extraction
python-docx Word document parsing
google-generativeai Gemini LLM API access
python-dotenv Environment variable management
pyttsx3 Text-to-speech conversion
fpdf Exporting summaries as PDFs
pandas Excel/CSV processing
Pillow Image handling
psycopg2 PostgreSQL database integration
SQLAlchemy Database ORM
unicodedata2 Text sanitization
mimetypes File type detection

📝 License MIT License — use freely with attribution

🧠 How It Works

  • 📁 Users upload multiple file types (PDF, DOCX, TXT, etc.).

  • 📄 Text is extracted and combined into a unified context with progress tracking.

  • 👤 Users can:

  • Ask natural language questions

  • Request summaries (full or topic-based)

  • Search for keywords

  • Download responses as PDF, text, or audio

  • View and export chat history

  • 🧠 Prompts are constructed using document context and sent to Gemini 2.0 Flash.

  • 📤 Responses are displayed with advanced UI styling and error handling.

  • 📥 Exports include detailed formatting and database storage.

📬 Contact

Built by Sanjai For suggestions or contributions, open an issue or pull request on GitHub: https://github.com/SANJAI-s0/IRS_genai/issues

🚀 Notes for Lighter Version

  • Features: Simplified UI with basic CSS, no progress bars or tooltips, and reduced error handling.

  • Database: Stores chat history in PostgreSQL with minimal configuration.

  • Limitations: Supports fewer file types (no PPTX or images) and lacks advanced UI features.

  • Recommendation: Use for quick prototyping or resource-constrained environments.

About

An Information Retrieval System is a system designed to search, retrieve, and present information from a large collection of unstructured or semi-structured data (like documents, web pages, articles) based on user queries.

Resources

Readme

License

MIT license
Activity

Stars

1 star

Watchers

1 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages